From: Jason Gunthorpe <jgg@mellanox.com>
To: "Weiny, Ira" <ira.weiny@intel.com>
Cc: Yuval Shaia <yuval.shaia@oracle.com>,
"dledford@redhat.com" <dledford@redhat.com>,
"leon@kernel.org" <leon@kernel.org>,
Moni Shoua <monis@mellanox.com>,
Parav Pandit <parav@mellanox.com>,
Daniel Jurgens <danielj@mellanox.com>,
"kamalheib1@gmail.com" <kamalheib1@gmail.com>,
Mark Zhang <markz@mellanox.com>,
"swise@opengridcomputing.com" <swise@opengridcomputing.com>,
"Berg, Johannes" <johannes.berg@intel.com>,
"willy@infradead.org" <willy@infradead.org>,
Michael Guralnik <michaelgur@mellanox.com>,
Mark Bloch <markb@mellanox.com>,
"dan.carpenter@oracle.com" <dan.carpenter@oracle.com>,
"bvanassche@acm.org" <bvanassche@acm.org>,
Max Gurtovoy <maxg@mellanox.com>,
Israel Rukshin <israelr@mellanox.com>,
"galpress@amazon.com" <galpress@amazon.com>,
Denis Drozdov <denisd@mellanox.com>,
Yuval Avnery <yuvalav@mellanox.com>,
"Dalessandro, Dennis" <dennis.dalessandro@intel.com>,
"will@kernel.org" <will@kernel.org>,
Erez Alfasi <ereza@mellanox.com>,
"linux-rdma@vger.kernel.org" <linux-rdma@vger.kernel.org>,
Shamir Rabinovitch <srabinov7@gmail.com>
Subject: Re: [PATCH v1 00/24] Shared PD and MR
Date: Fri, 23 Aug 2019 11:57:36 +0000 [thread overview]
Message-ID: <20190823115731.GA12847@mellanox.com> (raw)
In-Reply-To: <2807E5FD2F6FDA4886F6618EAC48510E898ADD18@CRSMSX102.amr.corp.intel.com>
On Thu, Aug 22, 2019 at 08:10:09PM +0000, Weiny, Ira wrote:
> > On Thu, Aug 22, 2019 at 09:58:42AM -0700, Ira Weiny wrote:
> >
> > > Add to your list "how does destruction of a MR in 1 process get
> > > communicated to the other?" Does the 2nd process just get failed WR's?
> >
> > IHMO a object that has been shared can no longer be asynchronously destroyed.
> > That is the whole point. A lkey/rkey # alone is inherently unsafe without also
> > holding a refcount on the MR.
> >
> > > I have some of the same concerns as Doug WRT memory sharing. FWIW I'm
> > > not sure that what SCM_RIGHTS is doing is safe or correct.
> > >
> > > For that work I'm really starting to think SCM_RIGHTS transfers should
> > > be blocked.
> >
> > That isn't possible, SCM_RIGHTS is just some special case, fork(), exec(), etc all
> > cause the same situation. Any solution that blocks those is a total non-starter.
>
> Right, except in the case of fork(), exec() all of the file system
> references which may be pinned also get copied.
And what happens one one child of the fork closes the reference, or
exec with CLOEXEC causes it to no inherent?
It completely breaks the unix model to tie something to a process not
to a FD.
> > Except for ODP, a MR doesn't reference the mm_struct. It references the pages.
> > It is not unlike a memfd.
>
> I'm thinking of the owner_mm... It is not like it is holding the
> entire process address space I know that. But it is holding onto
> memory which Process A allocated.
It only hold the mm for some statistics accounting, it is really just
holding pages outside the mm.
Jason
next prev parent reply other threads:[~2019-08-23 11:57 UTC|newest]
Thread overview: 49+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-08-21 14:21 [PATCH v1 00/24] Shared PD and MR Yuval Shaia
2019-08-21 14:21 ` [PATCH v1 01/24] RDMA/uverbs: uobj_get_obj_read should return the ib_uobject Yuval Shaia
2019-08-21 14:21 ` [PATCH v1 02/24] RDMA/uverbs: Delete the macro uobj_put_obj_read Yuval Shaia
2019-08-21 14:21 ` [PATCH v1 03/24] RDMA/nldev: ib_pd can be pointed by multiple ib_ucontext Yuval Shaia
2019-08-21 14:21 ` [PATCH v1 04/24] IB/{core,hw}: ib_pd should not have ib_uobject pointer Yuval Shaia
2019-08-21 14:21 ` [PATCH v1 05/24] IB/core: ib_uobject need HW object reference count Yuval Shaia
2019-08-21 14:53 ` Jason Gunthorpe
2019-08-27 16:28 ` Yuval Shaia
2019-08-27 18:18 ` Jason Gunthorpe
2019-08-21 14:21 ` [PATCH v1 06/24] IB/uverbs: Helper function to initialize ufile member of uverbs_attr_bundle Yuval Shaia
2019-08-21 14:21 ` [PATCH v1 07/24] IB/uverbs: Add context import lock/unlock helper Yuval Shaia
2019-08-21 14:57 ` Jason Gunthorpe
2019-08-21 14:21 ` [PATCH v1 08/24] IB/verbs: Prototype of HW object clone callback Yuval Shaia
2019-08-21 14:59 ` Jason Gunthorpe
2019-08-21 14:21 ` [PATCH v1 09/24] IB/core: Install clone ib_pd in device ops Yuval Shaia
2019-08-21 14:21 ` [PATCH v1 10/24] IB/mlx4: Add implementation of clone_pd callback Yuval Shaia
2019-08-21 14:59 ` Jason Gunthorpe
2019-08-21 14:21 ` [PATCH v1 11/24] IB/mlx5: " Yuval Shaia
2019-08-21 14:21 ` [PATCH v1 12/24] RDMA/rxe: " Yuval Shaia
2019-08-21 14:21 ` [PATCH v1 13/24] IB/uverbs: Add clone reference counting to ib_pd Yuval Shaia
2019-08-21 14:21 ` [PATCH v1 14/24] IB/uverbs: Add PD import verb Yuval Shaia
2019-08-21 15:00 ` Jason Gunthorpe
2019-08-21 14:21 ` [PATCH v1 15/24] IB/mlx4: Enable import from FD verb Yuval Shaia
2019-08-21 14:21 ` [PATCH v1 16/24] IB/mlx5: " Yuval Shaia
2019-08-21 14:21 ` [PATCH v1 17/24] RDMA/rxe: " Yuval Shaia
2019-08-21 14:21 ` [PATCH v1 18/24] IB/core: ib_mr should not have ib_uobject pointer Yuval Shaia
2019-08-21 14:21 ` [PATCH v1 19/24] IB/core: Install clone ib_mr in device ops Yuval Shaia
2019-08-21 14:21 ` [PATCH v1 20/24] IB/mlx4: Add implementation of clone_pd callback Yuval Shaia
2019-08-21 14:21 ` [PATCH v1 21/24] IB/mlx5: " Yuval Shaia
2019-08-21 14:21 ` [PATCH v1 22/24] RDMA/rxe: " Yuval Shaia
2019-08-21 14:21 ` [PATCH v1 23/24] IB/uverbs: Add clone reference counting to ib_mr Yuval Shaia
2019-08-21 14:21 ` [PATCH v1 24/24] IB/uverbs: Add MR import verb Yuval Shaia
2019-08-21 14:50 ` [PATCH v1 00/24] Shared PD and MR Jason Gunthorpe
2019-08-22 8:50 ` Yuval Shaia
2019-08-21 23:37 ` Ira Weiny
2019-08-22 8:41 ` Yuval Shaia
2019-08-22 14:15 ` Doug Ledford
2019-08-26 9:35 ` Yuval Shaia
2019-08-22 16:58 ` Ira Weiny
2019-08-22 17:03 ` Jason Gunthorpe
2019-08-22 20:10 ` Weiny, Ira
2019-08-23 11:57 ` Jason Gunthorpe [this message]
2019-08-23 21:33 ` Weiny, Ira
2019-08-26 10:58 ` Yuval Shaia
2019-08-26 10:29 ` Yuval Shaia
2019-08-26 12:26 ` Jason Gunthorpe
2019-08-26 9:51 ` Yuval Shaia
2019-08-26 10:04 ` Yuval Shaia
2019-08-26 10:10 ` Yuval Shaia
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20190823115731.GA12847@mellanox.com \
--to=jgg@mellanox.com \
--cc=bvanassche@acm.org \
--cc=dan.carpenter@oracle.com \
--cc=danielj@mellanox.com \
--cc=denisd@mellanox.com \
--cc=dennis.dalessandro@intel.com \
--cc=dledford@redhat.com \
--cc=ereza@mellanox.com \
--cc=galpress@amazon.com \
--cc=ira.weiny@intel.com \
--cc=israelr@mellanox.com \
--cc=johannes.berg@intel.com \
--cc=kamalheib1@gmail.com \
--cc=leon@kernel.org \
--cc=linux-rdma@vger.kernel.org \
--cc=markb@mellanox.com \
--cc=markz@mellanox.com \
--cc=maxg@mellanox.com \
--cc=michaelgur@mellanox.com \
--cc=monis@mellanox.com \
--cc=parav@mellanox.com \
--cc=srabinov7@gmail.com \
--cc=swise@opengridcomputing.com \
--cc=will@kernel.org \
--cc=willy@infradead.org \
--cc=yuval.shaia@oracle.com \
--cc=yuvalav@mellanox.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).